1 General Introduction

The Forward Proton Detector (FPD) is a set of 18 detectors along the beam line at both sides of the D0 colision point. The detectors will measure the protons and the antiprotons deviations from the original trajectories after elastic interactions at the "colision" point.

We will take measurements on the horizontal plane (10 detectors) and on the veretical plane (8 detectors). The deviation knowledge togheter with the knowledge of the magnetic field value from quadrupole and dipole magnets and the electric field from separators will give us the particle trajectory and important informations about the Physics of the interaction.

The detector is a scintillating fiber grid in three directions on three parallel planes. Each scintillating fiber is spliced to a clear optical fiber that ends on a multiple photomultiplier. Protons and antiprotons passing through the scintillating fiber generate light pulses that are guided to the photomultipliers and transformed to electric pulse signals.

The optical fiber groups end in plastic "cookies" and the cookies fit inside the photomultiplier sockets. There is a large central scitillating block for tigger purpouse that detects any particle crossing the active area of the detector. The light generated by that block is coupled to a light guide and it goes to the central photomultiplier.

The Roman Pot Castle is a mechanical device that will move the detectors on a perpendicular direction to the beam line. It will be a precise movement with a resolution about 10 microns controlled dy DO controls system. The detector works at atmospheric pressure and will be isolated from vaccum by a thin stainless steel window. The "Pot" is the device that holds and moves the thin window box with the inside detector.

Some questions we had to face: How to handle and to fit the detector inside the Pot? How to handle and support all photomultipliers? The cartridge idea!

The goals:

• It must holds and protects the detector and optical fibers.
• It must be light tight.
• It must hold all photomultipliers.
• It must provide a way to fit the detector inside the window box without stressing the plastic detector.
• It must be easy to replace the photomultipliers.
• It must be easy to assemble and to connect the optical fiber cookies.
• It must be easy removable. Detectors can not get warm during the
• It must have a size compatible with the floor hole.
• It must be easy removable. Detectors can not get warm during the baking process.
• It must have a easy lock mechanism to the Pot.
• It must be easily machined and use commom and commercial materials to be cheap.

The design and the prototype.

• The two rods will guide the detector until it reaches the guide pins.
• The 5/16" tube will pull the detector against the window box bottom.
• The soft silicone rubber interface will remove the stress on the plastic detector due to small cartridge misaligment.
• It will be a split type body and can be totaly deassembled.
• The photomultipliers will be held in place by using soft silicone rubber to make them selfaligmened into the sockets.
• The body will be made from 8" IPS sch. 80 commercial pipe.
• The male-female collar will lock the cartridge to the Pot.

Movement and controls.

Pot movement will be achieved by use of a screw type piston (2mm pitch), a large diameter nut, a worm gear reduction (120:1) and a stepping motor. (1.8 degree/step). Each motor axle half revolution corresponds to a 8.3 microns Pot displacement.

Constraint: the time spent to completly remove the Pot from the beam line can not be too large, tipicaly less then 3 minutes. That means a high speed motor.

Problems:

• Motor torque goes down quickly as the speed increases.
• It is not reliable to electronic count the electric pulses from the driver to determine the position at high speeds. Motor can loose steps.
• Modern inteligent stepping motor drivers are microprocessed and use some type of serial line to exchange information with the computer. But the distance is too large (300 ft or more) and we don't know how microprocessors survive in a radioactive enviroment.
• Long distances can cause problems with electric ground and optical link increases the cost and complexity.
• Optical encoders to determine motor axle position are easily damaged by radiation.

Solutions:

• Development of a custom magnetic encoder with Hall effect sensors to generate an electric pulse each motor axle half turn.
• Use the reliable "D0 Rack Monitor Module" near the Pot, in the Tev tunnel, to exchange informations with D0 online control system.
• Development of a custom "Index Interface" to translate the electric signals between the Rack Monitor and a "dumb" driver. The Index Interface must do the limit switches logic, must count the encoder pulses, must control the driver and generate alarm signals.
• LVDTs will be used as a backup position device.